The Last Season (9 page)

Clinton Hester, convinced that the Liberator had crashed somewhere in Kings Canyon, was determined to honor his son's service by bringing his body home. Year after year, he searched the high country himself, with any volunteers he could recruit. His quest was relentless, but after more than a decade of searching, he had not discovered a single clue to substantiate his theory.

In 1959, after fourteen years of methodically combing the mountains, Clinton Hester died from a heart ailment. One year after his death, in July 1960, his son's bomber was found by a ranger on a geological survey in the Black Divide range near LeConte Canyon. The plane had crashed into a 12,500-foot peak and exploded on impact. Some of the debris ended up in what would be known as Hester Lake. The elder Hester had come within a few miles of the wreckage.

The Hester Liberator was often referenced to illustrate just how overwhelming a search for a person in the High Sierra could be, considering that the wreckage from a 70-foot-long shiny silver bomber had eluded detection for a decade and a half.

Still, there was only one story in the parks' history of a missing person on foot who had not been found during the course of a search-and-rescue operation. His name was Fred Gist, a 66-year-old real estate appraiser from San Luis Obispo who had disappeared just beyond the southwest boundary of the Bench Lake ranger's patrol area on the Monarch Divide. Gist was last seen by his companions on August 19, 1975, near where Dougherty Creek flows into the crystalline waters of the Lake of the Fallen Moon.

Five search dogs and four trackers from the U.S. Border Patrol joined twenty-six rangers and volunteers from across the state on an intense, leave-no-stone-unturned search that began two days after Gist's disappearance. It was learned that Gist had been packed in with horses and wasn't a particularly strong hiker, so the search area was fairly compact, about three miles across. Using classic strategy of the time, it was grid-searched with dogs; according to the case incident report, “not a trace of the missing person was found.”

The search was called off on the seventh day, after high-resolution military photographs of the area produced no results. Fred Gist's fate had been a mystery for more than a decade until backpackers found his skull near Dougherty Creek. Without any knowledge of the mystery, they left the skull on the doorstep of the Simpson Meadow ranger station with a hand-drawn map showing where it had been found. Ironically, they had named the skull “Fred.”

The Gist search illustrates, perhaps more so than the Hester plane crash, how these mountains can hide a person, even with trained search teams combing an area. There was, however, one significant difference between Gist and Randy. Gist was wholly unprepared for the freezing nights; he reportedly didn't even have a sleeping bag with him.

Not only was Randy extremely fit, he also had with him survival gear and the knowledge to use it. He just had to hang on and let searchers find him somewhere in the 80-square-mile search area. No doubt about it, he was the classic needle in a massive haystack—but times had changed and so had modern-day search techniques.

In 1976, shortly after the Gist search, a lieutenant colonel in the U.S. Air Force named Robert Mattson came up with a brand-new method for prioritizing ground search areas. His innovative strategy, first published in the spring 1976 issue of
Search and Rescue Magazine,
came to be known as the Mattson method or the Mattson consensus. It was inspired by the pioneering work of B. O. Koopman, who, as a member of the U.S. Navy's Operations Evaluation Group, created a mathematical approach to locating enemy submarines in the vast oceans during World War II. So effective was the strategy, Koopman and his group
were credited with being key to winning the battle against German U-boats in the Atlantic.

The Mattson consensus has, for the most part, remained the favorite strategy of SAR professionals, including Coffman, who implemented its classic approach as the leader of the search effort at Bench Lake.

According to Mattson, experts who knew something about either the missing person or the terrain should be brought together; these individuals should be “the most informed and experienced personnel available.” In this case the rangers knew both Randy and the Sierra. After collecting as much information as possible about the victim and the area, Coffman divided the overall search area into reasonably sized segments. Then, using a secret ballot, each ranger assigned each segment a number value—high for areas where Randy most probably was, low for least-probable areas. According to Mattson, it was “best to do this privately because it will insure [sic] that even the meeker individuals will be able to express their opinion without being intimidated by the more vocal members of the group.”

Though Coffman ran the show and knew the history behind the theory, the rangers knew the drill and spoke the same acronym-heavy language. POA, for example, was “probability of area,” the probability that Randy was in a certain segment. ROW stood for the “rest of the world” and considered the possibility that Randy was somewhere other than inside the designated search area.

The percentage points assigned by each ranger for sixteen segments plus the theoretical ROW segment had to add up to 100 points. Nobody could assign a zero for any segment. That would mean they knew with certainty that Randy was not in that particular segment, which was impossible. In his article, Mattson had taunted readers for such optimism in the face of unknowns: “If you KNOW where the survivors are, why are you searching!!!!!???!!!!”

The overall message Mattson conveyed, above and beyond the mathematical approach, was “Never discard information, keep an open mind, use common sense, and dig, dig, dig for information.”

And dig they had. Coffman had a notebook full of notes to prove it.

In his logbook Randy had reported going south on the John Muir Trail to Pinchot Pass twice, once to the summit and the second time over the top to Woods Creek. Acting on their knowledge of Randy's habits as a ranger, they deduced that it was unlikely he'd gone that direction again—either via the JMT or any cross-country routes that eventually met up with it in that southerly direction.

On the other hand, Randy had not yet been to Lake Basin—which Durkee, Lyness, and Graban knew was a sacred place for him. Nor had he covered the cross-country routes in Upper Basin or any of the tucked-away gems north of the Bench Lake Trail, including Dumbbell Lakes and Marion Lake. Using this line of reasoning, the rangers threw out ideas of probable distances and places Randy might have visited on a three-to four-day patrol.

The information-gathering process had taken hours, but the voting process took about twenty minutes. Not surprisingly, the Lake Basin area (Segment F) was unanimously valued as the highest-percentage POA at 26.20 percent, while Marion Lake and its surrounding cirque (Segment G) was the second-most-probable consensus at 19.20 percent. The ROW option was voted as the lowest POA for everybody except Durkee, who assigned that option a curiously high percentage compared with the other rangers. The higher value prompted Coffman to ask, “You think Randy might have left the park? Why?”

“I told Coffman that Randy's life was in turmoil,” says Durkee, “though I didn't go into details with Lo sitting right there next to me.” Durkee also kept quiet about what he described as a “very slight, but unshakable” suspicion that his friend might have gone off to some special place and ended his life.

After Coffman dismissed them till morning and the other rangers had wandered off to their respective sleeping spots, Durkee made a discreet detour to the door of the station. Randy's note was still pinned to the canvas. The date he'd written was June 21. But it was July. Everyone else had discounted the mixup of “J” months as an honest slip of the pen, but Durkee couldn't stop thinking that it was a potential
clue to Randy's mindset at the time. He reprimanded himself for his paranoia and pushed aside the tent flap. As always, Randy's residence was spartan. “Randy never was much for putting up pictures or drapes to make his stations more homey,” says Durkee. “It was a minimalist base camp.”

Quickly, Durkee's headlamp beam found its mark: the steel footlocker where he knew Randy would have kept his sidearm. As expected, it was padlocked. He gave the lock a tug, just in case. Solid. He then turned his attention to Randy's military field desk—an olive drab rectangular wooden box with a leather handle on either end. The front was a row of drawers and cubbyholes topped by a worn-smooth working surface that folded up to reveal a storage compartment. Inside he found the expected stacks of mandatory reading: the new NPS 9 Law Enforcement Policy and Guidelines binder; a few inches' worth of backcountry policy, which Randy could recite from memory, mostly because nothing had changed much in the past decade; some EMT refresher manuals; a stack of citations; and the recently proposed, but not implemented, meadow management plan that had been dispersed during training to some of the backcountry rangers. Atop this particular document was a pen, and within the pages were Randy's notes and suggestions.

“It was a work in progress,” says Durkee, “which told me Randy intended to come back.” With that rationalization, the mentally exhausted ranger retired to his tent.

Coffman continued to plan into the night. Dave Ashe was his point man in the frontcountry, to whom he relayed—among other things—the results of the consensus.

Ashe and another ranger named Scott Wanek had organized an impromptu incident command post at the Kings Canyon fire station. They transformed a dormitory into a planning room and began the process of spreading the word to various state emergency response groups—a network of organizations that included the California Rescue Dog Association (CARDA) and volunteer SAR teams from different counties throughout the state. The military and state highway patrol were put
on standby, with potential requests for air support and personnel. The emphasis in requesting personnel was on expert hiker skills. Coffman had told Ashe to make it very clear: “The search area is complicated, dangerous, off-trail terrain.” Ashe, in turn, conveyed that he wanted “quality, not quantity.” The underlying message was “We don't want to have to rescue the rescuers.”

Somehow, even with all his other duties, Ashe found a few minutes to prep CASIE for data. CASIE, or Computer-Aided Search Information Exchange, was a program designed to simplify most of the calculations related to managing a search emergency using modern search theory and terminology. Once it has been plugged into CASIE, an overwhelming search area and operation becomes more easily digestible by date or segment. A glance at a computer printout provides basic information about the manner of searching a certain segment (air, foot, dog team, etc.) and how effective the searchers believe they were in “clearing” that area. Using this method to keep track of a large mass of land, the leader of the search—in this case, Coffman—would cross off search segments once he felt confident they were clear.

This, of course, presumes the missing person is not on the move and has not reentered an area already cleared, the reason for the “hug a tree” strategy preached at wilderness survival classes. Another difficulty is that segments are generally considered surface areas—not underwater, underground, or under a rock slide. In an area as vast as the Morgenson SAR, where only two segments were smaller than 500 acres, the majority were around 2,000 acres, and one segment was initially more than 7,000 acres, a thorough surface search was difficult enough. Compounding the challenge, the high country has a myriad of streams and rivers that empty into hundreds, if not thousands, of lakes. Nearly every peak has dozens of active rock-slide and snow avalanche paths, any of which could bury or otherwise conceal an injured or deceased victim.

Randy could be a few yards from a shouting search team, yet not be discovered. The same team could be employing an air-scent-trained search dog but if Randy was downwind, the dog would not catch his
scent. A search-dog handler described the nature of a SAR in the High Sierra as an “organized search in chaotic terrain.” That description didn't even begin to explore the depth of the chaos if one other possibility was included.

What if Randy didn't want to be found?

 

CHIEF RANGER DEBBIE BIRD
was saddle-sore and weary when she arrived at the Road's End trailhead just after dark. The horseback ride from Vidette Meadow, where she'd last seen ranger Lo Lyness, was about 16 miles. She'd seen the helicopter activity and even without a working radio had deduced that a search-and-rescue operation for Randy had been initiated.

She drove immediately to Cedar Grove, where she saw the fire station lit up and alive with activity. She looked for Coffman, figuring he'd be in charge of the search, but instead found Ashe, who brought her up to speed on the few details currently available. “Coffman,” he told her, “is at Bench Lake.”

This wasn't what Bird wanted to hear. She felt that Coffman and Ashe weren't thinking far enough ahead—not treating this as a major incident. She sensed the search would “evolve into something more,” she explains. “Once these things get going, it becomes very difficult to catch up logistically.” As the chief ranger, Bird could pull rank and issue an order to use an incident command system (ICS), but with Randy Coffman, “who is a, well…a very talented SAR ranger,” she says, “it was much more effective to work at making
your
idea become
his
idea, rather than issuing some kind of order.”

With this in mind, Bird radioed Coffman, whom she hoped would be receptive to an ICS.

The incident command system was developed in the 1970s in response to a series of fires in Southern California during which a number of municipal, state, federal, and county fire authorities collaborated to form FIRESCOPE (Firefighting Resources of California Organized for Potential Emergencies). During that firestorm, a lack of coordination and cooperation between various agencies had resulted
in overlapping efforts and, worse, major gaps in the response. Consequently, some areas were overstaffed with firefighters, while undermanned property nearby was destroyed. From this experience came the original ICS model for managing wildland fires and emergencies at the city, county, and, eventually, federal level. In 1985, the Park Service adopted and pioneered the ICS for search-and-rescue and other emergency operations and is widely credited with honing the system.